Hear roleplay replies in a matching synthetic voice.
Generates roleplay audio using a voice cloned from public video or designed from an uploaded image.
When to use it
Use when the user wants to converse with a real or fictional person, hear content in their voice, or speak with someone pictured.
Ask it to roleplay a named person or someone shown in an image; it returns generated audio with the response text.
Python 3 must be available to run the bundled extraction and voice-design scripts and the tts skill.
ffmpeg must be installed for extracting reference audio.
This skill
Noiz voice-design API
Uploads images to Noiz
YouTube or Bilibili
Searches public speaking videos